Monthly Archives

October 2007

Facebook now fourth largest site in the world

By | Facebook | 2 Comments

As per Comscore Facebook is now the fourth largest site in the world as measured by monthly page views.

Amongst the cognescenti the fashion du jour seems to be to talk Facebook down.  All the time I am hearing stories from people who are using Facebook less.  Yet the stats don’t lie (or at least not too much….).

I would explain this contradiction by saying that Facebook is like the footballs in this comment that Oli Barrett posted on Broadstuff:

Some sites are like Slinky Springs. Fun on Christmas Day, back in the box by New Year.

Others are more like footballs. You play with them like crazy over Christmas, then even though your usage drops, you still keep using it pretty regularly.

I am well through the ‘Christmas phase’ on Facebook and for me it is the status updates and Facebook applications that drive this regular usage.

Europe’s base of angel investors and serial entrepreneurs will grow quickly

By | Entrepreneurs, Venture Capital | One Comment

From the ever brilliant Marc Andreessen:

  • companies simply develop faster these days than they did in the past. Microsoft and Oracle, for example, both needed 10 years of incredibly work to get to their IPOs (both founded in ’76, IPO in ’86), and they only had a few hundred employees each when they went public — and those were the two biggest software successes of their era.

Versus these days, when many companies are founded, built, scaled up, and sold (or, yes, taken public!) in a few years.

  • Also because of the faster cycle time, when you start company #2 you can assume that it won’t necessarily consume the next 10-20-30 years of your life — you can probably build something successful over say 5 years, maybe 8 years max, and so you’re not committing the rest of your life.This makes it easier for people to say, OK, hey, it worked once, I’ll try it again.

This points are all about the entrepreneur side of the equation.  If it only takes half as long as previously to build a successful company you will get serial entreprenuers twice as fast.

The same is true for the investors in those businesses.  They will exit in half the time and will therefore be able to fund twice as many businesses (at least).

This is good news for us here in Europe where the start-up ecosystem is still relatively young.  The successes of recent years have given us a good number of serial entrepreneurs and angel investors who are making a big difference already – what you get from this class of people more than any other is the following three ‘C’s:

  1. Cash – for early stage companies
  2. Confidence – they show it can be done
  3. Companies – hopefully great ones!

If Marc is right, and I think he is, then the ranks of serial entrepreneurs and angels will swell more quickly than ever before.  Particularly as this is a great moment in time to be building companies.

To end on a note of hope, also from Marc:

For the first time in history, you have a global market of 1+ billion people, all connected over an interactive network where they’re all a click away from you. That’s amazing.

And 100 million new people are being added to that count every year, and that will continue for the next 30 years.

A huge and growing market makes all kinds of magical things possible, and I think that’s what we’re seeing now.

Flock 1.0 beta changing the game for social networks?

By | Facebook, Identity, Social networks, TechCrunch | 7 Comments

I have been using the Flock 1.0 beta for a few days now and the People Sidebar feature has really got me hooked. The product is well hyped (Techcrunch 40 winner) and has lots of great features (review) – but the People Sidebar could have a significance which goes beyond this latest skirmish in the browser wars.

Through tight integration with social media services the People Sidebar gives you an aggregated view of your network across the different services that you use. The picture below shows the top 3-4 people across my Twitter and Facebook networks.

This reduces the importance of having all my friends on the same network. It may even eliminate it.

As Facebook sweeps all before it people are struggling with compartmentalising their identity. It is problematic when your boss can see the side of you that is designed for your friends, or family. One solution to this problem is for Facebook et al to add features which allow you to show different parts of your personality to different friends – e.g. different groups can see different photos or status updates only go out to the portion of your network that you want to see them.

Flock is now giving us a glimpse of an alternative future. Instead of having your whole network in the same service and using tools within that network to compartmentalise your identity, you can leave the different parts of your network in the services that are most appropriate for them and manage them through an aggregator service in your browser.

This has profound implications for the value of social networks. If the approach Flock is pioneering wins out (and it is very early days yet, both in terms of number of services integrated and functionality available at the aggregate level) then I think we will see a world where there are multiple large social networks. Today it looks like there will only be space for one or two.

Friday fun – $15bn for Facebook, don’t believe the hype

By | Facebook, Microsoft, Social networks | No Comments

This is from the comments on a Fake Steve Jobs post about two hedge funds apparently investing $500m in Facebook at the same $15bn valuation Microsoft just paid.  (Thanks to Alan for the pointer.)

Picture this: Zuckerberg’s got a big fish on the line for real money at a $15 Bill valuation. But is having trouble landing it. So he calls up his buddy at MS (formerly of Groove Networks) and his other contacts from working with MS so far on selling ads. They work out a deal where they get to say publicly that MS has dropped $240 Mill for 1.6%. What isn’t so public is that a lot of it works out to funny-money; MS product, MS services contracts, consulting fees, etc. And a good deal of the real cash MS looks to get back with side contracts for future purchases from MS. So the real money spent is much smaller but Facebook still gets real investment because MS is just a ringer in the crowd to suck in the real marks. And MS gets their hedge on Facebook at a discount by trading on their name. Plus MS gets sales to Facebook on the books to add a little juice to their earnings number.

I don’t know about the conspiracy theory part, but I would be surprised if there isn’t some substance to the idea that Microsoft gets some kickbacks which they deducted from the headline $240m purchase price when doing their valuation calculations.

Musings on the future for virtual worlds

By | 3D, Business models, Second Life, Social networks, Virtual Worlds | 5 Comments

There are two big takeaways for me from the Virtual Worlds Forum conference I have been attending for the last couple of days.

The first is a firming up of a conclusion that I have been coming too for a while – there is huge opportunity at the nexus of virtual worlds, games (probably casual) and social networks.

Habbo Hotel is a good example of this, as are new companies that are here at the conference MoshiMonsters from Mind Candy and Moipal.

Games, social nets and virtual worlds are coming together and that is creating opportunities for new companies that are radically different from their forbears that have focused in one of the three areas. I think the best will be easy to learn, work well with short session times be social, and have an interesting and continually developing content story. The revenue model will be about subscriptions and/or virtual goods and/or advertising.

Which brings me to my second point – there is already huge opportunity (by which I mean dollar opportunity) for brand advertising in virtual worlds.
Until now I had seen the empty brand sponsored spaces in Second Life and the comparatively small numbers of active users and concluded that the brand budgets coming into the world are experimental, and were not at the point where they would really start to scale.

Not so.

The combination of deep engagement and response rates as high as 30% are getting brands to the point where they are parting with budgets they would normally reserve for campaigns with high headline impressions numbers. The key to this is the quality of the campaigns which contribute significant new content to the worlds. Habbo Hotel in particular is blazing a trail here.

World of Warcraft is the new golf

By | Networking, Virtual Worlds | 18 Comments

From Lord Putnam’s keynote address at the Virtual Worlds Forum in Kings Cross, London which is going on over the next couple of days:

“people are saying World of Warcraft is the new golf for technology focused networkers”

I love it. The parallels are legion – the addiction, the time away from the family, the incomprehensibility to non-players – and now people playing it as much to be social as for the game itself!

To generalise this – building something which is fun AND helps people with their social lives is a great way to get a really sticky community.

Emotional contagion in social networks

By | Consumer Internet, Facebook, Second Life, Social networks, Virtual Worlds | 2 Comments

In the 1960s William Condon extensively researched the non-verbal interaction which surrounds speech. Beyond the obvious body language of looking into people’s eyes and hand movements it turns out there is a world of tiny interactions that happen at incredibly high speed. He wrote:

Your body’s locked precisely with your speech. You can’t break out of this no matter what you do. Your eyes even blink in synchrony with your speech.” Movements appear to begin, change, or end on the same film frame that a new vowel or consonant begins – within about four-hundreths of a second in the new sound. “The synchrony of the listener with the speaker is just as good as my own synchrony with myself.” An auditory-motor reflex in the central nervous system might allow, even force, a listener’s movements to synchronize with a speaker’s voice far faster than any conscious reaction time. “We’re almost in auditory touch. When I speak to you, my thoughts are translated into muscle movements an and then into airways that hit your ear, and your eardrum starts to oscillate in absolute synchrony with my voice. In essence there’s no vacuum between us – it takes only a few milliseconds for a sound to register in the brain stem, 14 milliseconds for it to reach the left hemisphere.

This goes a long way to explaining the power of face to face meetings compared with phone calls and online interaction. I found my way to it through Gladwell’s Tipping Point.
It also highlights the importance of building conduits for expressing and sharing emotions into social networks.

For those that can be bothered to get over the hump of learning to use it properly SL does this really well – and the power of the emotions that people feel through their avatars is pretty well documented now – remember the avatars rioting over copybot? This is possible because there is a large range of tools available for people to signal emotions to one another in Second Life – including expressions (XYZ smiles, or blushes, or shrugs etc. etc.), the way someone dresses, the use of IM rather instead of public speech, giving of gifts, dancing, all the way to user programmed interactions.

Facebook also does this well, albeit at a much lower level of intensity suitable for a less immersive social experience. The tools there include pokes, writing on walls, status updates, private messages and virtual gifts. Thanks to JP for this insight.

In a related point, World of Warcraft is powerful because of the emotions evoked when teams pull together to do missions. It is incredible hearing people talk about the way they go about this, and how exciting it is. The exhilaration of a team victory is familiar to most of us and the widespread use of Skype for real time chat as teams undertake missions is important for getting the emotions across.
Regular readers of this blog know of my enthusiasm for virtual worlds and social media. What is dawning on me is the importance of thinking explicitly about the emotional aspects of any given service. In particular what the emotions will be and how they will be communicated.

This point is clearly not relevant for all internet services (e.g. delicious) but there are elements of it in most of the successful ones. It is particularly true in any 3D/virtual worlds play – where the whole point of being 3D is that it makes the experience imersive and more conducive to the transfer of emotions.

Seedcamp download – some good advice on how to think about product

By | Consumer Internet, Entrepreneurs, Seedcamp | No Comments

The following post has been sitting in draft form since Seedcamp week in early September.  I have tried to pick out the key insights from a couple of panels/presentations where people discussed product – they were mostly thinking about consumer internet.

These points are in no particular order.

  1. If you are building something you would use yourself you have a greater chance of success.  This is an oft talked about point, but one which really came home to me when I heard Michael Birch describe how he started Bebo.
  2. Use your own product – obsessively.  Also known as ‘dogfooding’ (i.e. eat you own…..).  That is how you will iron out all the quirks and get to the best user experience.  One CEO described how his team hates it when he goes to conferences because they know that he will use his downtime there to obsessively test their site.  They are big on this idea at Google.
  3. Find the ‘nub’ of your product and only build stuff that fits with that.  You should be able to capture the ‘nub’ in a single sentence.  If you find yourself wanting to build stuff that doesn’t fit with the nub then it is probably time to re-examine it….
  4. Don’t forget the customer – an old chestnut, but easily done
  5. Iterate obsessively – all the successful internet entrepreneurs I know are big on this point.  Continuously try new things on your site and see how they impact your key metrics.  Keep the good, throw out the bad and then try the next thing.  Over and over again.  The best way to do this is to set up your site so you can A-B test – i.e. let 5% of the traffic see the new feature and then move it to the main site only if it works.
  6. Know what you aren’t doing – this is the key to all strategy, and something that people often find hard to grasp, in my experience anyway.
  7. Lot’s of startup success comes from random groups of people unexpectedly getting hold of a product and running with it.  Be live to that possibility and nurture it if it starts to happen.  This happened at Skype where the countries which took off were not the ones they expected.
  8. Understand the metrics for your site, look for the bottlenecks (or key drop off points) and work on them – even if they are the most difficult, or even seemingly outside your control.  If you have a major drop off point on your customer conversion path all the tweaking in the world elsewhere won’t make much difference until you fix the big one.
  9. You need to start with vision and passion but at some point switch to data driven analysis of customer and traffic data.  Judging the point at which to switch is tricky.

The bull case for Facebook

By | Consumer Internet, Facebook, Social networks | No Comments

Forget $10bn – Lee Lorenzen thinks Facebook will be worth $100bn.  Whilst I wouldn’t go that far his arguments are not entirely crackers – although they do have one flaw that I will come to at the end of this post.

Lee Lorenzen has set up a venture fund, Altura Ventures, where the “whole portfolio will be made up of companies that have a shot of leveraging the dominance of Facebook”.

His main point is that Facebook can be as big as Microsoft.

The comparison should be with Microsoft [not Google] in terms of the lock-in that can be achieved when you control both the operating system and most of the killer apps on the platform.  ….
With Facebook, their lock-in will come from consumers who can’t easily decouple themselves from Facebook without giving up their network of friends and with application developers who will have to write for Facebook to reach a critical mass of connected consumers who trust their friends invitations/newsfeeds more than any 3rd party advertsing pitch  ……
Facebook’s entire business is based on bringing larger number of REAL PEOPLE closer to one another than has ever been possible before and keeping them informed about and engaged in one another’s business and lives.

But, as Lee concedes, Facebook’s ability to monetise it’s customer base is unproven.  They are reportedly making profits of $30m on revenues of $150m at the moment, whilst Microsoft is making profits of $14bn on revenues of $50bn.

I am with Lee in believing that Facebook has legs as a business, and the points he makes are good in that regard, but when you look at the Microsoft results alongside Facebook’s it becomes crystal clear just how much they need a killer revenue model.

This mooted deal between Microsoft and Facebook and the $10bn valuation associated with it has me wondering if they have a magic solution to this problem in the pipeline somewhere, but for now it really isn’t clear what that might be.  As I have been reading about CPMs and behavioural targeting over the last couple of days I have started thinking that Facebook will need something more than this if it is to live up to expectations.

Don’t get me wrong, Facebook will still be a great business, in fact it already is.  All I am talking about here is whether it goes on to be a genuine internet giant.

More page views means lower CPMs

By | Advertising, Consumer Internet, Social networks | 6 Comments

Up until now I have thought that CPMs on social networks are low mostly because people are familiar with the layout of pages and their eyes immediately go to the content that they want.  In other words ad-blindness is particularly acute.

I still think that is a large part of the explanation, but the fact that click through rates fall with the number of page impressions has also got to be important.

It turns out that the longer someone has been on a site the less likely they are to click on an advert.  The relationship might look something like this:

These numbers are hypothetical, but they imply that someone who has looked at 10 pages on a site is 1/6th as likely to click on an ad as someone who is still on their first page (rough reading of the graph as 3% clickrate for one ad-view and 0.5% clickrate for 10 ad-views).

Given that CPMs are generally calculated as a function of click through rate the implication is that CPMs on a site where the average user views 10 pages should be 1/6th of the CPMs on a site where the average user views only one page.  (The maths here is a little off, but you get my drift.)

That leaves us with the paradox that if you increase the average number of page views per user your CPM will fall.

As we know social networks are sites where average page views per user are typically very high – this argument goes some way to explaining their low CPMs.  It also rings true intuitively – if someone is busy clicking through 150 pages in a Facebook session she is most unlikely to stop and click on an ad.
This post is largely taken from 5 things that make your social network monetise like crap by Andrew Chen.  He makes a number of other interesting points in the post which is well worth a full read.  I will leave you with one other – there will be a range of CPMs on any given site, but people tend to quote the highest.